Acetylene Hydratase from Pelobacter acetylenicus : functional Studies on a Gas-Processing Tungsten, Iron-Sulfur Enzyme by Site Directed Mutagenesis and Crystallography

(1)

Acetylene Hydratase from Pelobacter acetylenicus

Functional Studies on a Gas-Processing Tungsten, Iron-Sulfur Enzyme

by Site Directed Mutagenesis and Crystallography

Dissertation

zur Erlangung des akademischen Grades eines Doktors der Naturwissenschaften

Dr. rer. nat.

vorgelegt von

Dipl.-Biol. Felix ten Brink

Konstanz, April 2010

1. Referent: Prof. Dr. P.M.H. Kroneck

(2)

(3)

Table of contents

Zusammenfassung

P. acetylenicus ist ein strikt anaerobes, mesophiles Bakterium, das auf Acetylen als einziger Kohlenstoff- und Energiequelle wachsen kann (Schink, 1985). Im ersten Schritt der Vergärung wird ein Molekül H2O an die Kohlenstoff-Kohlenstoff-Dreifachbindung des Gases Acetylen addiert und Ethenol gebildet, welches zu Acetaldehyd umlagert. Diese Reaktion wird von dem W,FeS Enzym Acetylen Hydratase (AH) katalysiert. Zwar liegt die dreidimensionale Struktur des Enzyms bei hoher Auflösung (1.26 Å) vor (Seiffert et al., 2007), dennoch blieben viel Fragen bezüglich des mechanistischen Ablaufs der Reaktion offen. In dieser Arbeit wurde die Addition von Wasser an Acetylen im Aktivzentrum in zwei unterschiedlichen Ansätzen untersucht, um sie im Detail zu verstehen.

1. Reaktion in kristallo

Kristalle der AH(W) aus P. acetylenicus wurden unter Überdruck mit Acetylen oder Kohlenmonoxid inkubiert, um die Bindung des Substrats oder des Inhibitors CO an das Aktivzentrum mittels Röntgenstrukturanalyse aufzuklären. In weiteren Experimenten wurden Kristalle der AH mit dem Endprodukt der enzymatischen Reaktion, Acetaldehyd, oder mit dem strukturell verwandten Molekül Propargylalkohol (C3H4O) versetzt.

Insgesamt wurden über 100 Kristalle unter Variation des Gasdrucks und der Reaktionsdauer umgesetzt. Hieraus wurden mehrere gute Datensätze erhalten, die zu erfolgreichen Strukturanalysen führten. In zwei Strukturen konnten Positionen von möglichen Acetylenmolekülen auf Grund von Änderungen in der Elektronendichte entdeckt werden. In beiden Fällen waren die Dichten von zwei Wassermolekühlen so eng zusammen gerückt, dass sie eine hantelförmige Dichte bildeten. Eine Position war am Rand des Proteins zu finden, in einer Tasche nahe des Eingangs des Substrat-Kanals. Die andere Position eines möglichen Acetylenmoleküls war in der Nähe des Wolframatoms am Aktivenzentrum, aber

(8)

hergestellt, in der das Wolfram während der Anzucht der Bakterien gegen Molybdän ausgetauscht wurde (AH(Mo)). Da die AH(Mo) nicht unter den gleichen Bedingungen gute Kristalle bildete, wie das entsprechende Wolfram-Enzym, musste nach neuen Bedingungen gesucht werden. Ein Ansatz erbrachte kleine nadelförmige Kristalle der AH(Mo), die bisher noch nicht die erhofften Beugungsdaten lieferten. Diese Versuche zur Kristallisation der AH(Mo) werden momentan weitergeführt.

2. Ortsgerichtete Mutagenese von Aminosäureresten im Aktivzentrum

Um Aminosäuren, die für die Reaktion der AH wichtig sind, mit ortsgerichteter Mutagenese zu identifizieren, wurde ein System zur heterologen Expression der AH in E. coli entwickelt. Die Aktivität des exprimierten Enzyms war im Vergleich zur nativen AH aus P. acetylenicus, deutlich niedriger. Als Hauptursachen konnten der Einbau des Molybdänkofaktors und des Metalls während der Expression der AH ausgemacht werden. Das Hinzufügen einer N-terminalen Chaperon- Bindesequenz der Nitratreduktase NarG aus E. coli war hilfreich sowohl in Bezug auf den Einbau des Metalls als auch des Kofaktors. Das Ergebnis ist ein rekombinantes Fusions-Protein, das über eine Aktivität von 80% von der des nativen Enzymes verfügt.

Die ortsgerichtete Mutagenese zeigte die Wichtigkeit der Carboxylsäuregruppe im Rest Asp13 für die Reaktion der AH. Der Austausch anderer Aminosäuren zeigten, dass der konservierte Rest Lys48 dabei hilft das Metal für die Reaktion zu aktivieren, in dem er den reduzierten [4Fe-4S]-Zentrum mit dem Q-MGD verbindet, auch wenn die Reaktion der AH keinen netto Elektronentransfer beinhaltet,. Mit dem Austausch von Ile142 konnte die Substrat-Bindestelle der AH identifiziert werden. Ile142 ist Teil eines hydrophoben Ringes aus sechs Aminosäuren, der eine Tasche am Ende des Substrat-Kanals bildet. Ein Acetylenmolekül in dieser Tasche befindet sich Direkt über dem Sauerstoff-Ligand des Wolframatoms und Asp13.

(9)

Summary

P. acetylenicus is a strictly anaerobic, mesophilic bacterium, that is able to grow on acetylene as sole source of carbon and energy (Schink, 1985). The first step in the fermentation of acetylene is the hydration to form acetaldehyde. This reaction is catalyzed by the W/FeS enzyme acetylene hydratase. While the three-dimensional structure of this enzyme has been solved at high resolution (Seiffert et al., 2007), the molecular basics of the reaction mechanism remained unclear. In this work, the mechanism behind the hydration of a gaseous substrate by a metal-dependent enzyme was studied in two approaches.

1. Reactivity in crystallo

Crystals of AH(W) isolated from P. acetylenicus were incubated under elevated pressure of acetylene or CO to solve the X-ray structure with the substrate or an inhibitor at the active site. Other crystals were soaked with the end product of the enzymatic reaction, acetaldehyde, or with a structural analog of acetylene, propargylalcohol (C3H4O) to get a three dimensional structure with one of these compounds at the active site.

Several data sets from over 100 experiments under varying conditions of gas pressure and reaction time could be collected and the structures were solved. In two structures locations of putative acetylene molecules could be detected. One was located at the outskirts of the protein in a pocket near the entrance of the substrate channel. The other putative acetylene was in close proximity to the tungsten atom at the active site, but with a distance of about 5 Å too far away for the hydration to take place.

For further crystallographic experiments a variant of AH, where the tungsten was replaced by molybdenum during cultivation of the bacteria, was prepared. Since this AH(Mo) did not crystallize under the same conditions as AH(W), screens for

(10)

To identify amino acid residues, which are important for the reaction of AH, by site directed mutagenesis, a system for the heterologous expression of AH in E. coli was developed. Activity of the expressed AH was significantly lower compared to the native enzyme isolated from P. acetylenicus. Molybdenum cofactor and metal insertion were identified as the main limiting factors during the expression of AH.

Addition of an N-terminal chaperone binding sequence from the E. coli nitrate reductase NarG helped to increase both, cofactor and metal insertion, resulting in a recombinant fusion protein that exhibited 80% of the activity of the native enzyme.

Site directed mutagenesis demonstrated the importance of the carboxylic acid group in the residue Asp13 for the reaction of AH. Further amino acid exchanges indicated that, although the reaction of AH does not involve a net electron transfer, the conserved residue Lys48 helps to activate the metal site for the reaction by linking the reduced [4fe-4S] cluster and the Q-MGD of the molybdenum cofactor.

By exchange of Ile142 the substrate binding site of AH could be identified. Ile 142 is part of a hydrophobic ring of six amino acids that forms a pocket at the end of the substrate channel. An acetylene molecule positioned in this pocket is located directly above the oxygen ligand of the tungsten atom and Asp13.

(11)

1. Introduction

1.1 Molybdenum and tungsten in biological systems

Chemical properties and bioavailability of molybdenum and tungsten

Molybdenum and tungsten are the only elements of the second (Mo) and third (W) row of transition metals with known biological function (Hille, 2002). Due to the lanthanide contraction, which takes place before tungsten in the periodic table, the atomic radii as well as the electron affinity of molybdenum and tungsten are virtually the same. Although both elements have a broad range of possible oxidation states (-II to +VI) only the +IV, +V and +VI oxidation states seem to be biologically relevant (Kletzin and Adams, 1996). To be biologically active molybdenum and tungsten must be bound to a special cofactor called molybdopterin. This cofactor is present in all molybdenum and tungsten enzymes with the exception of nitrogenase where the molybdenum ion is bound in a large molybdenum iron sulfur cluster (Schwarz et al., 2009).

Molybdenum and tungsten both have an abundance of 1.2 ppm in the earth crust (Greenwood and Earnshaw, 1990). Molybdenum is mainly present as insoluble molybdenite (Mo^IVS2) but the oxidation state +VI is also quite common as Me²⁺MoO42-. Tungsten is mainly present as tungstate (Me²⁺WO42-) because tungstenite is readily transformed to WO42- and H2S (Kletzin and Adams, 1996) (eq. I).

(eq.I) WS₂ + 4H₂O WO₄^2- + 2H⁺+ 2H₂S + 2[H]

For uptake and use of molybdenum and tungsten all organisms rely on the soluble anions

(12)

Molybdenum or tungsten enzymes are found in nearly all organisms, with Saccharomyces as a prominent eukaryotic exception (Schwarz et al., 2009). Plants depend on the molybdenum nitrate reductase and on the iron, molybdenum nitrogenase, found in bacteria associated to their roots, for their nitrogen supply (Mendel and Bittner, 2006). Humans and other mammals need molybdenum in enzymes, such as sulfite reductase and xanthine dehydrogenase. Bacteria carry a wide variety of molybdenum enzymes like nitrate reductase, formate dehydrogenase, dimethyl sulfoxide reductase or trimethylamine N- oxide reductase.

In contrast to molybdenum enzymes, tungsten enzymes are only found in thermophilic and hyperthermophilic bacteria and archaea with the exception of some anaerobic mesophiles like Pelobacter acetylenicus (Hille, 2002). In those extremophiles tungsten enzymes catalyze the same reactions as molybdenum enzymes in other organisms (Stiefel, 1997).

Deep sea hydrothermal vents, where many tungsten dependent organisms are found, reflect the condition on the primitive earth (Schwarz et al., 2009). Transition metals like nickel and tungsten, and sulfur compounds may have played an important role in building efficient early life catalysts (Kroneck, 2005). When the dioxygen concentration in the atmosphere raised, molybdenum became bioavailable in form of molybdate and replaced tungsten.

This theory is supported by the fact that some enzymes are still active when molybdenum is replaced by tungsten or vice versa, as well documented for dimethyl sulfoxide (DMSO) reductase (Sigel and Sigel, 2002). In many cases the enzyme activity is influenced by the metal exchange. Although the DMSO reductase from Rhodobacter capsulatus is a molybdenum enzyme, its tungsten form shows an increased rate for the DMSO reduction but a decreased rate for the DMS oxidation (Stewart et al., 2000). The trimethylamine N- oxide reductase from E. coli is a molybdenum enzyme as well, its tungsten form has a lower turnover rate for TMAO. Interestingly, the tungsten TMAO reductase reduces not only TMAO, but also DMSO and related compounds (Buc et al., 1999). In the case of tungsten acetylene hydratase from P. acetylenicus the activity of the molybdenum form is reduced (Abt, 2001).

(13)

1. Introduction

On the other hand many bacteria and their enzymes are very selective in metal incorporation. In Desulfovibrio alaskensis molybdenum or tungsten are inserted into the formate dehydrogenase (FDH), depending on the concentration in the medium, but only molybdenum is inserted into the aldehyde reductase (Andrade et al., 2000). The FDH from Desulfovibrio gigas and Syntrophobacter fumaroxidans incorporates only tungsten, despite of suitable molybdenum concentrations in the media (Moura et al., 2004). This was also the case in Pyrococcus furiosus, where only tungsten was incorporated into the AOR family enzymes (Sigel and Sigel, 2002).

In view of these findings, bioavailability doesn’t seem to be the only important parameter, which determines the metal that is incorporated in the active site. Studies on molybdopterin model complexes showed that the redox potential of the molybdenum or tungsten complex may be another factor that influences the metal requirement of certain enzymes (Schulzke, 2005). The difference in the redox potential for the transition M^IV M^V is only 30 mV at 25°C, with the potential for the Mo complex being slightly more positive than that of the W complex. At temperatures above 70°C, however, where the tungsten using extremophiles are living, the potential of the W complex becomes more positive than that of the Mo complex (Schulzke, 2005). Therefore, a change in the redox potential of the active site metal may be the explanation why the metal exchange influences the activity of the enzyme and why some bacteria are highly selective towards metal uptake and incorporation.

Molybdenum and tungsten enzymes

Until recently over 50 enzymes containing a mononuclear molybdenum or tungsten center have been identified (Hille et al., 1998). Despite the similarity between the chemical properties of molybdenum and tungsten, the enzymes containing tungsten represent only a small percentage compared to the molybdenum enzymes (Johnson et al., 1996).

(14)

to a pterin, forming the molybdenum cofactor (Moco, Fig. 1-1) which is the active compound at the catalytic site of all Mo and W enzymes (Brondino et al., 2006). An exception of this rule is the nitrogenase, where the molybdenum is bound by a special MoFe-cluster.

Figure 1-1: Structure of the pterin cofactor associated with molybdenum and tungsten containing enzymes. In prokaryotic enzymes, the cofactor exist as the dinucleotide of guanine, adenine, cytidine or hypoxanthine (Hille et al., 1998).

Molybdenum cofactor and enzyme classification

Molybdenum cofactor (molybdopterin or moco) containing enzymes are found in all three kingdoms of life. Eukaryotic molybdenum enzymes usually contain only one pterin ring as cofactor, whereas prokaryotic enzymes can contain different cofactors consisting of one or two pterin rings and a nucleotide covalently linked to the pterin moiety (Fig. 1-2). The nucleotide can be one of the four bases guanine, adenine, cytidine or hypoxanthine. (Hille, 2002; Moura et al., 2004). The effects of the nucleotide on the chemical properties of the molybdenum or tungsten ion complex are not yet fully understood. The main function of the cofactor is to position the molybdenum or tungsten ion in the active site, to control the redox properties of the metal and to shuffle electrons from and to the metal via the pterin rings. Additionally, the pterin system with its different conformations and redox states could help to transfer electrons to other cofactors or prosthetic groups of the enzyme (Kisker et al., 1997).

(15)

1. Introduction

HN

N N

H HN

O

O SH

SH

OPO₃^2- H₂N

1 2 3

4 5

6 8 7 10

9 1'

2' 3'

4'

HN

N N

H HN

O

O SH

SH

O H₂N

1 2 3

4 5

6 8 7 10

9 1'

2' 3'

4'

P O₂^2-

O P O₂^2-

O

NH N

N N

O

HO OH

NH₂ (A)

(B)

(C)

NH H N

N

NH O

O S

S

O NH₂

O₂^2- P O O₂^2- P O HN

N N O N

O OH HO

H₂N

HN

N N

H HN

O

O S

S

H₂N O P

O₂^2- O P

O₂^2- O

NH N N

N O

O

HO OH

NH₂ Mo O

Q-MGD

P-MGD

Figure 1-2: Cofactors of molybdenum and tungsten enzymes.

(A) Molybdenum cofactor = molybdopterin = moco. The tricyclic form was observed in all crystal structures of enzymes containing this cofactor (Kisker et al., 1998).

(B) Molybdopterin guanosine dinucleotide (MGD) form as found in some bacterial

enzymes (Stiefel, 1997).

(C) Extended molybdenum cofactor (bis-MGD) as found in arsenite oxidase from Alcaligenes faecali (Ellis et al., 2001).

Molybdenum and tungsten enzymes are a heterogeneous group of enzymes. In addition to the different forms of moco they may also contain other redox cofactors like iron sulfur

(16)

divided into four families: sulfite oxidase, xanthine oxidase, dimethyl sulfoxide (DMSO) reductase and aldehyde / ferredoxin oxidoreductase family (Hille, 2002; Moura et al., 2004) (Fig. 1-3). The central metal ion is typically complexed by the dithiolene sulfur of the moco and by a variable number of oxygen (oxo, hydroxo, water, O-Ser, O-Asp), sulfur (C-Cys, sulfido) and selenium ligands (Se-Cys, selenido) (Fig. 1-3). Of all four families, enzymes of the DMSO reductase family show the widest diversity of active site structures.

Molybdenum Enzymes

Tungsten Enzymes

Figure 1-3: Active site structures of the molybdenum and tungsten enzyme families based on their three dimensional structures. A) Molybdenum enzymes of the DMSO reductase family. X: O-Ser in DMSO reductase, transhydroxylase; Se-Cys in formate dehydrogenase; S-Cys in periplasmic nitrate reductase; O-Asp in respiratory nitrate reductase; B) Tungsten enzymes of the DMSO reductase family including acetylene hydratase (S-Cys) and formate dehydrogenase (Se-Cys) (Hille, 2002; Moura et al., 2004).

(17)

1. Introduction

Heterologous expression of molybdopterin-dependent enzymes

Heterologous expression of molybdenum and tungsten enzymes still represents a highly complex problem. The molybdenum cofactor must be incorporated into the protein at the right moment of biosynthesis to ensure the correct folding of the protein. Additionally, many molybdenum and tungsten enzymes contain not only moco but also other redox cofactors, which must be incorporated into the protein as well.

For expression of enzymes from the sulfite oxidase and xanthine oxidase families, a special E. coli strain called TP1000 (Palmer et al., 1996) has been used. This strain has a kanamycin cassette inserted into the mobAB gene region. Since the mobAB genes encode the first enzymes in the pathway from molybdopterin to bis-molybdopterin guanosine dinucleotide (bis-MGD), molybdopterin is not further processed and accumulates in the E.

coli TP1000 cells. Therefore this strain is very useful for the expression of simple molybdenum and tungsten enzymes, which contain the basic (non-modified) molybdopterin cofactor (sulfite and xanthine oxidase families). Using this system several enzymes from both families have been successfully expressed heterologously in E. coli:

human sulfite oxidase (Temple et al., 2000), Rhodobacter capsulatus xanthine dehydrogenase (Leimkuhler et al., 2003), Thermus thermophilus sulfite oxidase (Di Salle et al., 2006) and mouse aldehyde oxidase (Schumann et al., 2009).

Additionally, several enzymes of these families have been expressed without the TP1000 system. The rat sulfite oxidase was expressed using E. coli JM109 (Garrett and Rajagopalan, 1994), while the retinal aldehyde oxidases from rabbit and mouse were expressed in E. coli BL21 (Huang et al., 1999).

Much less was done in the case of enzymes from the DMSO reductase family to which acetylene hydratase (AH) belongs. The earliest attempts have been described by Bilous and Weiner (1988) who succeeded to overexpress DMSO reductase in E. coli (Bilous and Weiner, 1988). Later, The biotin sulfoxide reductase and the DMSO reductase from Rhodobacter sphaeroides f. sp denitrificans were expressed heterologously in E. coli

(18)

The molybdenum cofactor is synthesized in a multi-enzyme cascade (Fig. 1-4) (Schwarz et al., 2009). Starting from GTP, the pterin ring system is formed first. The pterin is then sulfurated, resulting in the metal-binding pterin with a copper atom bound to the dithiolene sulfur atoms. The copper atom is then replaced by molybdenum to form the molybdenum cofactor (moco).

Figure 1-4: Biosynthesis of molybdenum cofactor (Schwarz et al., 2009). The first four steps, leading from GTP to moco, are present in all organisms. The fifth step, resulting in bis-MGD is only present in prokaryotes. Enzymes from E. coli are named in the ‘Mo’- nomenclature.

(19)

1. Introduction

In prokaryotes, a second multi-enzyme cascade processes moco further to form the bis- MGD or MCD forms of the cofactor (Fig. 1-5) (Mendel, 2005).

Figure 1-5: Enzymes involved in the further processing of moco in prokaryotes (Mendel, 2005).

In E. coli the biosynthesis of moco is regulated by three factors (Anderson et al., 2000): (i) molybdate. If molybdate is present in the cell, the molybdate binding protein ModE will bind it and then activate the operons involved in moco synthesis; (ii) dioxygen. The operons for moco biosynthesis are activated in the absence of dioxygen by the FNR oxygen sensor system; (iii) free active moco. Like in the case of many operons for chemically demanding synthesis of complex cofactors, the moco operon RNA has a riboswitch in front of the transcription starting point (Regulski et al., 2008). The RNA motif will bind free moco present in the cell. This will induce a conformational change in the three dimensional-structure of the RNA and block the ribosome binding and therefore the synthesis of the enzymes involved in moco biosynthesis.

(20)

Chaperone mediated cofactor delivery

Studies on the maturation of molybdenum cofactor containing enzymes have revealed a family of chaperones that facilitate the incorporation of moco into the apoenzyme during protein biosynthesis and prevent the export of periplasmic enzymes before moco is incorporated (Ilbert et al., 2004; Palmer et al., 1996; Pommier et al., 1998; Sargent, 2007).

Three member of this TorD family are known in E. coli: TorD for the TMAO reductase, DmsD for the DMSO reductase and NarJ for the Nitrate reductase (Sargent, 2007).

Complementation studies indicated that these chaperones are highly specific for their partner and cannot complement the absence of another chaperone (Ilbert et al., 2004).

The best studied chaperone is TorD. It acts in two ways: (i) It binds at the N-terminal TAT signal sequence of the moco containing TorA TMAO reductase. This way it keeps the protein unfolded until the cofactor is delivered and prevents the export of the protein without moco (Sargent, 2007). (ii) It binds to a yet unidentified second binding site where it facilitates the insertion of moco in the folding enzyme (Jack et al., 2004) (Fig.1-6).

Figure 1-6: Chaperone mediated cofactor delivery and protein export (Jack et al., 2004). The chaperone (in the case of TMAO reductase TorD) binds at the N-terminal TAT signal sequence and a second yet unidentified binding site. There it prevents the folding and export without the cofactor.

(21)

1. Introduction

Cytoplasmic enzymes in E. coli like the moco containing subunit of the nitrate reductase NarG have a modified TAT signal sequence where the twin arginine motif is missing.

Therefore the TAT exporters do not recognize the protein for export and it remains in the cytoplasm (Sargent, 2007). TAT signal sequences without the twin arginine motif have been identified at the N-terminus of many cytoplasmic moco enzymes in different bacteria like Bacillus subtilis or Paracoccus pantotrophus which indicates that the use of these sequences as chaperone binding site for moco insertion is widely spread (Sargent, 2007).

1.2 Iron sulfur clusters

Iron sulfur clusters are one of the most ancient ubiquitous, structurally and functionally diverse class of biological prosthetic groups (Beinert et al., 1997). In many molybdopterin containing enzymes like acetylene hydratase, transhydroxylase or xanthine oxidase iron sulfur cluster were also found.

In the simplest case the iron atom has a tetrahedrally coordinated site with four cysteine sulfurs, whereas in the more complex forms several iron atoms are bridged by inorganic sulfide (S^2-), the so-called acid labile sulfur. The most common types of iron sulfur centers comprise [2Fe-2S], [3Fe-4S] and [4Fe-4S] clusters with cysteine residues serving as terminal ligand of each iron atom (Fig. 1-7).

Rubredoxin

[2Fe-2S] Ferredoxin

[2Fe-2S] Rieske center

[3Fe-4S]

[4Fe-4S]

(22)

The genes for iron sulfur cluster assembly are widely conserved in all three kingdoms of life (Bandyopadhyay et al., 2008). Three distinct types of biosynthetic machineries emerged in bacterial, archaeal and eukaryotic organelles: NIF, ISC and SUF. The NIF system is involved in Fe-S cluster synthesis for proteins in nitrogen fixation. The ISC is used for general Fe-S cluster synthesis in bacteria and eukaryotic organelles. The third bacterial system plays a similar role as the ISC system, but is only active under iron limitation and oxidative stress (Bandyopadhyay et al., 2008).

The main function of the Fe-S cluster assembly systems is (i) to mobilize Fe and S from their storage sources, (ii) to assemble them to a Fe-S form and transport them and (iii) to transfer them to their final destination protein (Fontecave, 2006). In all three systems the clusters are assembled in so-called U-Type scaffold proteins (IscU, SufU or NifU) as [2Fe- 2S] clusters. [4Fe-4S] are built from two [2Fe-2S] clusters. The complete Fe-S clusters are then incorporated into the apo Fe-S protein (Fig. 1-8) (Bandyopadhyay et al., 2008).

Figure 1-8: Hypothesis for the mechanism of the IscS mediated [2Fe-2S]²⁺ and [4Fe-4S]²⁺ cluster assembly on IscU and transfer to the apo forms of the acceptor proteins (Bandyopadhyay et al., 2008).

(23)

1. Introduction

Studies on the assembly of ferredoxins and other Fe-S proteins revealed that the formation of Fe-S clusters was increased under respiratory growth compared to fermentative growth (Nakamura et al., 1999). Under anoxic conditions glycerol can be used as carbon source for respiratory growth of E. coli (Bilous and Weiner, 1988; Nakamura et al., 1999). The glycerol dehydrogenase, which catalyzes the first reaction in the reductive branch of the E.

coli glycerol fermentation, is coenzyme B12 dependent. In the absence of coenzyme B12 the bacteria are forced into respiratory growth and therefore to activate the Fe-S cluster assembly machinery (Lengeler et al., 1999). This may help to increase the Fe-S cluster incorporation into heterologously expressed proteins in E. coli.

1.3 Acetylene

Acetylene (C2H2) is a highly flammable gas that forms explosive mixtures with air over a wide concentration range (2.4 – 83 % Vol., material safety datasheet, Air Liquide GmbH, Germany). Set under pressure it can polymerize exothermically. It’s solubility of 45.5 mM in water is rather high compared to other gases like H2, N2 or O2 which have solubilities around 1 mM (Hyman and Arp, 1988). Therefore bacteria can easily live on the dissolved amounts of acetylene, if an adequate source is available. The first publication on the utilization of acetylene by bacteria appeared 75 years ago (Birch-Hirschfeld, 1932). Later, Norcadia rhodochorous was described to use C2H2 as sole source of carbon and energy in the presence of dioxygen (Kanner and Bartha, 1979). In 1980 an acetylene hydratase was found in the cell free extract of Rhodococcus A1, that grew on acetylene by anaerobic fermentation (de Bont and Peck, 1980). Like the acetylene hydrates from P. acetylenicus, this enzyme was reported to form acetaldehyde from water and acetylene and to be inhibited by dioxygen. The first stable pure culture of acetylene fermenting anaerobes was obtained by enrichment with acetylene from freshwater and marine sediments (Schink, 1985). Acetylene fermenting bacteria from estuarine sediments were again isolated by

(24)

Reactions of Acetylene

The carbon-carbon triple bond of the acetylene molecule consists of a σ-bond and two orthogonal π-bonds (Fig. 1-9). The bond distance is 1.20 Å.

1^st π bond in x, y plane 2^nd π bond in x, z plane Cylindrically symmetrical

set of π electrons

Figure 1-9: π-orbitals of acetylene

The hydrogen atoms of alkynes are relatively acidic compared to hydrogen atoms of alkenes or alkanes. Acetylene itself has a pKa of about 24 compared to 44 for ethylene (Hyman and Arp, 1988). The chemistry of acetylene is rather rich and diverse. A number of different reactions, such as reduction and oxidation, as well as electrophilic and nucleophilic additions do exist (Yurkanis-Bruice, 2004). Elecrophilic addition on alkynes tend to be much slower compared to the additions on alkenes, while nucleophilic additions to alkynes are much faster than those on alkenes (Bohlmann, 1957).

In biological systems acetylene is well known as inhibitor of microbial processes via interaction with the active sites of several metalloenzymes, such as nitrogenase, hydrogenase, ammonia monooxygenase, methane monooxygenase, assimilatory nitrate reductase, or nitrous oxide reductase (Hyman and Arp, 1988). Acetylene has been employed for the quantification of several important biological processes. For instance, the reduction of acetylene by nitrogenase serves as a measure for nitrogen fixation (Stewart et al., 1967), and the inhibition of N2O reductase is used to quantify denitrification (Rosner and Schink, 1995).

(25)

1. Introduction

Sources and bio-availability

Today acetylene is only a minor trace gas in Earth’s atmosphere. Mixing levels usually span 0.02 – 0.08 ppbv, depending on were the samples were collected (Oremland and Voytek, 2008). According to literature, the larger part of atmospheric acetylene is of anthropologic origin. Exhaust from combustion engine seem to be the main source of acetylene today (Whitby and Altwicker, 1978).

While acetylene is quite rare on earth, it can also be found among other prebiotic molecules in interstellar gas clouds (Thaddeus, 2006). Interestingly, a place where acetylene is more abundant than on earth, is Saturn’s moon Titan. Titan’s atmosphere is considered to be a cold model of earth’s early atmosphere 4 billion years ago (Oremland and Voytek, 2008). Photochemical processes in Titan’s upper atmosphere create acetylene, which sediments down to the surface (Schulze-Makuch and Grinspoon, 2005) (Fig. 1-10).

Figure 1-10: Proposed environmental condition on Titan. Photochemical reactions produce acetylene in the atmosphere. Due to the high specific gravity acetylene will sediment to the surface and accumulate along with other hydrocarbons (Schulze-Makuch and Grinspoon, 2005).

(26)

hydrocarbons (76-79% ethane, 7-9% propane, 5-10% methane, 2-3% hydrogen cyanide, 1% butene, 1% butane and 1 % acetylene) on Titans surface (Cordier et al., 2009; Tokano, 2009). Similar processes in earth’s early atmosphere and volcanic eruptions may have provided the developing life with acetylene as an easy accessible source of carbon and energy (Oremland and Voytek, 2008).

1.4 Acetylene Hydratase from Pelobacter acetylenicus

Pelobacter acetylenicus is a strict anaerobic, mesophilic bacterium that is able to grow with acetylene as sole source of carbon and energy. The first step in the metabolism of acetylene, the hydration of acetylene to acetaldehyde, is catalyzed by the enzyme acetylene hydratase (AH) (Rosner and Schink, 1995). The growth of P. acetylenicus on acetylene and acetylene hydrates activity depends on the presence of tungstate or molybdate in the growth medium (Rosner and Schink, 1995).

According to its amino acid sequence AH belongs to the DMSO reductase family. AH is a monomer with a molecular mass of 81.9 kDa (amino acid sequence) versus 73 kDa by SDS-Page (Meckenstock et al., 1999; Rosner and Schink, 1995). Pure AH from P.

acetylenicus isolated in the absence of dioxygen contained 4.4 ± 0.4 mol Fe; 3.9 ± 0.4 mol S; 0.5 ± 0.1 mol W and 1.3 ± 0.1 mol Molybdopterin guanine dinucleotide per mol enzyme (Meckenstock et al., 1999) . The activity of AH depends on the redox potential of the solution, with 50% maximum activity at –340 ± 20 mV (Meckenstock et al., 1999).

Therefore a strong reductant like Na-dithionite or Ti(III)-citrate has to be added to the enzyme assay. This finding is supported by biomimetic studies that could demonstrate the likely participation of a W(IV) site in the catalysis of the hydration of acetylene, while the W(VI) compound remained inactive (Yadav et al., 1997).

(27)

1. Introduction

Overall structure

A high resolution 3D structure of AH was achieved by X-ray diffraction analysis (Seiffert et al., 2007). The overall structure shows the typical four-domain fold of the DMSO reductase family members. The first domain contains the [4Fe-4S] cluster. Domains II-IV bind the molybdopterin guanine dinucleotide (Fig. 1-11).

Figure 1-11: Overall structure of acetylene hydratase from P. acetylenicus. The four fold domains, typical for DMSO reductase family members, are labeled in I) white; II) olive; III) red and IV) green. The region colored in gray as differently arranged than in other enzymes of the DMSO reductase family and closes the substrate channel typically found at this place (Seiffert, 2007).

Unique in the structure is the location of the substrate channel: instead of being located between domain II and III, as found in other DMSO reductase family enzymes, the substrate channel is located at the intersection of the domains I, II and III.

(28)

Active site

The putative active site of AH is located at the W atom in the center of the enzyme (Seiffert et al., 2007). The W atom is octahedrally coordinated by the four dithiolene sulfurs, a sulfur from the Cys141 residue and a sixth oxygen ligand (Fig. 1-12). From the crystal structure (1.26 Å resolution) it is not clear whether this ligand is a water molecule or a hydroxo ligand, since the oxygen tungsten distance of 2.04 is between the values expected for a H2O molecule (1.9 – 2.1 Å) or a –OH group (2.0 – 2.3 Å). Asp13 forms a hydrogen bond to this oxygen ligand. Next to Asp13 is Cys12, one of the four cysteines coordinating the [4Fe-4S] cluster. The whole arrangement of the active site is shielded towards the substrate channel by a ring of 6 large, hydrophobic residues (Ile14, Ile113, 142, Trp179, Trp293 and Trp472).

Figure 1-12: Active site arrangement in acetylene hydratase from P. acetylenicus. A) Residues of the active site. From the Oxygen tungsten distance in the crystal data it is not possible to discriminate whether the oxygen ligand is a water molecule or a hydroxo ligand B) Ring of 6 bulky, hydrophobic residues, shielding the active site towards the substrate channel (Seiffert et al., 2007).

(29)

1. Introduction

Reaction mechanism

A crucial question for the understanding of the detailed reaction mechanism of acetylene hydration by AH concerns the nature of the oxygen ligand bound to the tungsten atom (Seiffert et al., 2007). A putative acetylene binding pocket is located directly above this ligand (Fig.1-13) (Seiffert et al., 2007).

Figure 1-13: Putative acetylene binding pocket at the active site of AH. The C2H2 molecule (green) was modeled in the space above the oxygen ligand (Seiffert et al., 2007).

Two different mechanisms for the hydration of acetylene have been postulated, depending on the type of the oxygen ligand (Seiffert et al., 2007): a hydroxo ligand would be a strong nucleophile and would yield a vinyl anion with acetylene of sufficient basicity to deprotonate Asp13 and form the corresponding vinyl alcohol. Another water molecule could then bind to tungsten and becomes deprotonated by the basic Asp13, thereby regenerating the hydroxo ligand for the next reaction cycle. Alternatively, a bound H2O molecule would get a partially positive net charge through the proximity of the protonated Asp13, making it an electrophile that in turn could directly attack the triple bond in a Markovnikov-type addition with a vinyl cation as intermediate. In this model Asp13 remains protonated. To discriminate between these two possibilities further studies are necessary (Seiffert et al., 2007)

(30)

1.5 Scope of the study

As a member of the DMSO reductase family acetylene hydratase belongs to a group of enzymes, typically involved in redox reactions. Furthermore, AH harbors two MGD ligands bound to tungsten and a [4Fe-4S] cluster, two cofactors normally used for catalysis of redox reactions and electron transfer (Meckenstock et al., 1999). However, the hydration of acetylene to acetaldehyde does not involve a net electron transfer. This raises the question why a typical redox enzyme is used to mediate the hydration of acetylene.

Although the X-ray structure of AH gave a detailed view of the enzyme and it’s active site (Seiffert et al., 2007), the molecular basics of the reaction mechanism remained unknown.

Two different approaches were chosen to gain a deeper insight into the reaction mechanism of AH.

In the first approach crystals of AH(W) isolated from P. acetylenicus were incubated with acetylene and carbon monoxide to obtain a crystal structure with the substrate or an inhibitor bound at the active site. Additionally, crystals were soaked with acetaldehyde or propargylalcohol to solve the structure with the product of the enzymatic reaction or a structure with a derivative of acetylene at the active site.

Recent experiments with radioactively labeled acetylene demonstrated that AH(Mo) has a higher affinity towards the substrate than AH(W) (Fig. 1-14) (Seiffert, 2007). Since incubation of crystals of AH(Mo) promised a better chance to get a structure with the substrate bound, screening for crystal conditions of AH(Mo) was started.

1 2 3 4

Figure 1-14: Phosphor imager picture of AH incubated with ¹⁴C2H2. Lane 1+2: AH(W) as isolated and reduced; Lane 3+4: AH(Mo) as isolated and reduced; (Seiffert, 2007).

73 kDa

(31)

1. Introduction

In the second approach amino acids with putative functions in the proposed reaction mechanism were exchanged by site directed mutagenesis. Since genetic tools for site directed mutagenesis in P. acetylenicus were not available, a system for heterologous expression of AH in E. coli was established. With this system active AH and variants could be expressed in E. coli in sufficient amounts to clarify the role of certain amino acid residues in the reaction mechanism.

(32)

(33)

2. Materials and Methods

2.1 Chemicals and biochemicals

If not further specified chemicals were obtained in p.a. quality and were use without additional purification.

Buffers

Fluka: MES (2-(N-morpholino)ethane sulfonic acid); Merk. K2HPO4, NaHCO3, Na-acetate (trihydrate); Sigma: Na-cacodylate (trihydrate); Riedel-de-Haën: KH2PO4, Na2CO3; Roth:

Hepes (N-[2-hydroxyethyl]piperazine-N’-[ethane sulfonic acid]), Tris (tris- (hydroxylmethyl)-aminomethane);

Chromatographic resins

GE Healthcare: Chelating Sepharose Fast Flow, Resource15Q, Resource30Q; SuperDex^TM 200 HiLoad^TM 26/60;

Crystallization factorials

Crystal screen solutions (The Classics, The Classics Lite, The MPDs and PACT) were obtained from NeXtal Biotechnologies (now Qiagen).

Fluka: MPD (2-methyl-2,4-pentanediol, Ultra), PEG (Polyethylene glycol, BioUltra, molecular masses. 400; 1’000, 2’000, 4’000 and 8’000)

Dyes

Serva: bromphenol blue (sodium salt), Coomassie brilliant blue G-250 Gases

(34)

AppliChem: Guanidine HCl; Fluka: Na2WO4, Propargylalcohol; Merck: K3[Fe(CN)6], NaOH; 37.5% HCl, CoCl2·2H2O; MgCl2·6H2O, Na2MoO4; Dimethylsulfoxid (DMSO) Riedel-de-Haën: Acetaldehyde, Na2S2O4, Mg-acetate (tetrahydrate), Na-acetate, Na-nitrate, CuSO4·5H2O; Glycerol; Sigma: BCA (bicinchoninic acid solution); Roth:

Ethylenediamine-tetraacetate·2Na·2H2O (EDTA), Glycine, NADH (nicotinamide adenine dinucleotide) Rotiphorese^® Gel 30: 30% (w/v) Acrylamide with 0.8% (w/v) Bisacrylamide.

Proteins and enzymes

BioRad: low range SDS/PAGE molecular weight standards; Fluka: DNase I (desoxyribonuclease I); Serva: BSA (bovine serum albumin); Sigma: Alcohol Dehydrogenase from Yeast, Xanthine Dehydrogenase from milk.

Titanium(III) citrate was synthesized as described (Zehnder and Wuhrmann, 1976).

2.2 Acetylene Hydratase from P. acetylenicus

Cultivation of P. acetylenicus

Batch cultures of Pelobacter acetylenicus strain WoAcy1 (DSMZ 3246) were grown in freshwater medium (Rosner and Schink, 1995) at 30°C. The medium was sterilized at 121°C, cooled under N2/CO2 atmosphere (80%/20%, v/v), buffered with 30 mM NaHCO3

and reduced with 1 mM Na2S. After addition of vitamin solution (Widdel and Pfennig, 1981) and a modified trace element solution SL10 (Widdel and Pfennig, 1981) with the adequate concentrations of MoO42- and WO42- the pH was a adjusted to 7.0 – 7.4. Cultures were inoculated by 10% (by vol.) of a glycerol stock culture. Acetylene was provided at a concentration of 7-10% (respective to the medium volume) in the gas phase of the glass bottle.

(35)

Growth was monitored photometrically at 578 nm. The pH of the medium was maintained at 7.0 by addition of 1 M Na2CO3.

Cultures were harvested in the late exponential growth phase with an ultrafiltration cell (0,22 µm PVDF membrane, Millipore) and subsequent centrifugation at 10,000 x g for 35 min at 4°C. The resulting cell pellet was frozen in liquid N2 and stored at -70°C until further use.

AH(W)

In order to obtain the AH(W) from P. acetylenicus, cells were grown in the presence of 800 nM WO42- and 6 nM MoO42-. 20 l batch cultures of tungstate grown cells were harvested after 4-5 days at OD578= 0,8.

AH(Mo)

In order to obtain the AH(Mo), tungstate grown cells were transferred at least 6 times in medium containing 2 µM MoO42- and 2 nM WO42- before they were used to inoculate a 20 l batch culture. Cultures of molybdate grown cells were harvested after 6-7 days at OD578= 0.8.

Purification

AH(W) and AH(Mo) were purified, as described before (Abt, 2001), in the absence of dioxygen under a N2/H2 (94%/6%, v/v) atmosphere in an anaerobe chamber (Coy).

Chromatography was performed at 20°C in the anaerobe chamber on a FPLC system (GE Healthcare) equipped with a SPD-M10Avp Diode Array Detector (Shimadzu).

Frozen cells were thawed in the anoxic chamber and suspended in 50 mM Tris pH 7.5 supplemented with 1 mM PMSF, 10 mM MgCl2 and DNaseI. To prepare the crude extract,

(36)

ultracentrifugation (Optima LE-80K ultracentrifuge with a Ti45 rotor; Beckman) at 100,000 x g. The soluble fraction was treated by two consecutive ammonium sulfate precipitations (2.0 M and 3.2 M (NH4)2SO4). After each step precipitated protein was collected by centrifugation at 10000 x g for 30 min at 4°C. The pellet of the second precipitation step was suspended in 50 mM Tris pH 7.5, desalted by ultrafiltration and loaded on a Resource 30Q anion exchange column (GE Healthcare). AH containing fractions were identified with SDS-Page. The pooled fractions were diluted 2:1 with 50 mM Tris pH 7.5 and loaded on a Resource 15Q anion exchange column (Amersham). AH containing fractions were identified by SDS-Page, pooled and concentrated in a stirring cell (30 kDa cutoff regenerated cellulose membrane, Millipore) for loading on a SuperDex 200 gelfiltration column (Pharmacia). AH containing fractions were again identified with SDS-Page, pooled, concentrated to 10 mg/ml, frozen in liquid N2 and stored at -70°C until further use.

Crystallization

All crystallization experiments were done under exclusion of dioxygen under a N2/H2

(94%/6%, v/v) atmosphere in an anoxic chamber (Coy). Crystal plates were stored inside the chamber at 20°C. Sitting drop crystallization was done on Cryschem Plates (Hampton Research). For hanging drop crystallization Easy Xtal Tools (Qiagen) were used.

Crystallization of AH(W)

Crystals of AH(W) were prepared as described before (Seiffert et al., 2007). Crystals of AH(W) grew within 1-3 weeks from a 10 mg/ml protein solution in 5mM Hepes/NaOH pH 7.5 reduced by 5 mM Na2S2O4. Buffer exchange was done on a NAP5 column (GE Healthcare) followed by a concentrating step in a Vivaspin 500 centrifugal device (30 kDa cut off, Sartorius Stedim Biotech GmBH). 2 µl protein solution were mixed with 2.2 µl reservoir solution (0.1 M Na cacodylate, 0.2 M Mg acetate, 21% polyethylene glycol 8000, 0.04 M NaN3).

(37)

Incubation of crystals with C2H2 and CO

To get a structure with the substrate or an inhibitor bound at the active site, crystals of AH(W) were incubated with C2H2 in a Xenon Chamber (Hampton Research) or with either C2H2 at 0.75 bar or CO at 1.0 bar in a custom made chamber at the MPI for Biophysics at Frankfurt (Germany). Crystals were transferred in a cryoprotectant solution made from the reservoir by addition of 20% (v/v) methyl-2,4-pentanediol (MPD) and then incubated with C2H2 or CO in the chamber for 5-30 min. In a second approach crystals were first incubated with gas and then quickly soaked in the cryoprotectant. In both cases, the crystals were frozen and stored in liquid nitrogen.

Figure 2-1: Pressure cells for incubation of protein crystals with gases. Left: Xenon Chamber used for the first incubation experiments. Right: pressure cell from the MPI for Biophysics at Frankfurt (Germany). The Crystals were mounted with the cryo loop into the cell, set under elevated pressure of C2H2 or CO and then quickly frozen in liquid nitrogen.

(38)

Soaking of crystals with acetaldehyde and propargylalcohol

A second approach towards a crystal structure that might provide use with more information on the reaction mechanism of AH(W) was to soak crystals of AH(W) with propargylalcohol, or with acetaldehyde, the product of the enzymatic reaction.

Propargylalcohol (2-propyn-1-ol) is structurally related to acetylene. Earlier EPR experiments showed that propargylalcohol is one of the rare compounds that interacts with the metal sites of AH (Seiffert, 2007). Crystals of AH(W) were first transferred in a cryoprotectant solution made from the reservoir by addition of 20% (v/v) MPD. The crystals were subsequently transferred into cryoprotectant solutions with increasing amounts (50, 100, 250 and 500 µM) of acetaldehyde or propargylalcohol. When 500 µM acetaldehyde or propargylalcohol was reached the crystals were incubated for 5-30 min and then quickly frozen in liquid nitrogen.

H C C H H C C

OH C

O H H₃C

C O

A B C D

Figure 2-2: Compounds for incubation of crystals. A) acetylene; B) propargylalcohol; C) acetaldehyde; D) carbon monoxide

Crystallization of AH(Mo)

Initial screening for crystallization conditions of AH(Mo) was done with four NeXtal suite screens (Classics, Classics Lite, PACT, MPD) using both the sitting and hanging drop vapor diffusion method. All screen solutions were supplemented with 40 mM NaN3 and stored in the anoxic chamber for at least 3 days to remove oxygen. The protein was prepared in 5 mM Hepes/NaOH pH 7.5 reduced with 5 mM Na2S2O4 at concentrations between 7.5 and 12.5 mg/ml. The concentration was adjusted in a Vivaspin 500 centrifugal device (30 kDa cut off, Sartorius Stedim Biotech GmBH) after buffer exchange on a NAP5 column (GE Healthcare).

(39)

The crystal plates were stored in the anoxic chamber for 5-6 weeks. During this time period crystal formation was regularly checked. Solutions in which crystals grew were modified in respect of salt and precipitant concentration and buffer pH to optimize crystal growth. Grown crystals were transferred into a cryoprotectant solution made from the reservoir solution by addition of 20% (v/v) MPD and frozen in liquid nitrogen for diffraction experiments.

Data collection

Diffraction data were collected using synchrotron radiation (6.0-17.5 keV, 2.07-0.71 Å) at the X06DA (PX III) beamline at the Swiss Light Source (Paul Scherrer Institute, Villingen, Switzerland). The beamline was equipped with a mar225mosaic CCD detector.

Structure determination and refinement

All data sets were indexed, integrated and scaled using the Program XDS (Kabsch, 1993).

Initial molecular replacement and refinement were done with Phaser (McCoy et al., 2007) and RefMac5 (Murshudov et al., 1997) from the CCP4i suite (CCP4, 1994; Potterton et al., 2003), using the known structure of AH(W) (PDB ID: 2e7z ) (Seiffert et al., 2007) as template. Additional model building steps were done using COOT (Emsley and Cowtan, 2004) followed by refinement with RefMac5.

Graphical representation

Illustrations of protein structures were prepared with COOT (Emsley and Cowtan, 2004) and PyMol (Delano, 2002)

(40)

2.3 Heterologous Expression of AH in E. coli

Cloning

Cloning of AH into the pET24a(+) plasmid (Novagen) was done by Trenzyme GmbH (Konstanz).

Genomic DNA of P. acetylenicus was isolated from 1.5 ml of a dense culture using the DNA Blood and Tissue Kit (Qiagen). Genomic DNA and vectors were then sent to Trenzyme GmbH. There the AH gene was amplified by PCR and cloned into the vector using the NheI / XhoI restriction sites. A thrombin cleavage site was inserted in front of the C-terminal His-tag of the vector. Insertion of the gene was checked by restriction with NheI / XhoI and EcoRV / FspI and by sequencing of the insert. The resulting vector was named pET24_AH.

Transformation

Transformation of E. coli cells was done as described; Buffers and competent cells were prepared following the given protocol (Inoue et al., 1990). For propagation of plasmids E.

coli JM109 (Stratagene) was used. Plasmids from 2 ml over night cultures of E. coli JM109 were isolated using either the GeneElute Plasmid Miniprep Kit (Sigma) or the QIAprep Spin Miniprep Kit (Qiagen). Isolated plasmids were transformed into E. coli BL21 (DE3), E. coli BL21 (DE3) pLysS (Stratagene) or E. coli Rosetta (DE3) (Novagene) for test expressions of AH.

(41)

Test expressions

Expression of AH was tested either aerobically in 100 ml DYT medium (per liter: 16 g Trypton, 10 g yeast extract, 5g NaCl), or anaerobically in 100 ml mineral medium (Tab.

2.1). Both media were supplemented with adequate antibiotics: kanamycine for the pET24 vector and additional chloramphenicol for E. coli Rosetta (DE3)

Expression in aerobic cultures was induced at OD600 = 0.6 by addition of 250 µM IPTG.

Cells were harvested (10,000 x g, 30 min, 4°C) after 6 h at 25 °C and disrupted by three passages through a French Press at 110 MPa.

The mineral medium for anaerobic test expressions was sterilized at 121°C and cooled under N2 / CO2 (80/20% v/v). Afterwards 1 mM Na2S, a carbon source (0.4 % glucose or 0.5% glycerol), an e^- acceptor (50 mM Na2 fumarate, 100 mM NaNO3 or 70 mM DMSO), 10 µM Na2WO4 and a modified trace element solution SL10 (Tab. 2.2) was added.

Cultures were inoculated with over night grown 2 ml DYT cultures. Expression of AH was induced with 100 µM IPTG at OD600 = 0.9. After 24 h at 25°C expression the cells were harvested by centrifugation (10,000 x g, 30 min, 4°C) and disrupted by three passages through a French Press (110 MPa). Soluble and membrane fractions of aerobically or anaerobically grown cells were separated by ultra centrifugation (100,000 x g, 1:30 h, 4°C). Formation of AH and its solubility was checked by SDS-Page.

Compound [mM] g/l K2HPO4 x 3 H2O 60 13.75

KH2PO4 40 5.44

NH4Cl 10 0.53

MgCl2 x 6 H2O 2 0.41 Protein Hydro Lysate -- 0.5

Table 2-1: Mineral medium for anaerobic E. coli cultivation.

(42)

Compound [mM] mg/l FeCl2 x 4 H2O 7.5 1500 CaCl2 x 2 H2O 6.8 1000 MnCl2 x 4 H2O 10.0 2000

ZnCl2 0.5 70

CoCl2 x 2 H2O 0.8 130 CuCl2 x 2 H2O 0.01 2 NiCl2 x 6 H2O 0.1 24 H3BO3 0.1 6

Table 2-2: Modified trace element solution SL10 (Widdel and Pfennig, 1981) The compounds were dissolved in 10 ml of 25% HCl, the volume was adjusted to 1l and the solution sterilized at 121°C.

Addition of chaperone binding sites

To improve the insertion of molybdenum cofactor during protein biosynthesis, the N- terminal chaperone binding sites of E. coli TMAO Reductase (TorA) and Nitrate Reductase (NarG) (Sargent, 2007) were cloned in frame into the AH expression vector.

Genomic DNA from E. coli JM109 and pET24_AH vector was given to Trenzym GmbH (Konstanz). There the first 117 bp from the TorA gene and 108 bp from the NarG were amplified by PCR. The vector was opened at the NheI / NdeI restriction sites and the sequences of the chaperone binding sites were inserted in frame in front of the AH gene.

Insertion was checked by sequencing. The resulting vectors were named pET24_TorA-AH and pET24_NarG-AH.

(43)

Site directed mutagenesis

Site directed mutagenesis was done by PCR. The vectors pET24_AH or pET24_NarG-AH isolated from 2 ml over night cultures of E .coli JM109 were used as templates. Mismatch primers for the desired amino acids exchanges (Tab. 2-3) were ordered at Microsynth (Lustenau, Austria). The PCR reaction mix containing 0.05 Units/µl DNA polymerase (High Fidelity PCR Enzyme Mix, Fermentas), 0.2 mM deoxonucleotides (dNTP Bundle, Jena Bioscience), 10x High Fidelity PCR Buffer (Fermentas), 2 mM MgCl2, 1 µM Primer I, 1 µM Primer II and 0.02 ng/µl Template was put in a Mastercycler gradient thermocycler (Eppendorf). A PCR program (Tab. 2-4) was run over night. The samples of each amino acid exchange were spread over the whole gradient of annealing temperatures.

Amino acid

exchange Primer Sequence Length % GC Melting

point Sense 5’-GTCAGTCCTGCGCCATTAATTGTGTTGTAGAGGCTGAAG-3’ 39 bp 48.72 80.95°C Asp13_Ala

Antisense 5’-CAACACAATTAATGGCGCAGGACTGACAGACAACGTGC-3’ 38 bp 50.00 80.47°C Sense 5’-GTCAGTCCTGCGAGATTAATTGTGTTGTAGAGGCTGAAGTG-3’ 41 bp 46.34 79.28°C Asp13_Glu

Antisense 5’-CAACACAATTAATCTCGCAGGACTGACAGACAACGTGCTTC-3’ 41 bp 46.34 79.78°C Sense 5’-GTATTTGTATGGCGTCGGTGAATGCGGACACGATC-3’ 35 bp 51.43 81.02°C Lys48_Ala

Antisense 5’-GCATTCACCGACGCCATACAAATACTATTGGGGGGAGTTG-3’ 40 bp 50.00 81.62°C Sense 5’-GCCATGTATATGAGTATCGGGAATACAGCCGGAGTTCATAG-3’ 41 bp 46.34 79.00°C Cys141_Ser

Antisense 5’-GTATTCCCGATACTCATATACATGGCGGAAGTCCAGTTCGG-3’ 41 bp 48.78 79.16°C Sense 5’-CCATGTATATGTGTGCCGGGAATACAGCCGGAGTTCATAG-3’ 40 bp 50.00 81.83°C Ile142_Ala

Antisense 5’-GCTGTATTCCCGGCACACATATACATGGCGGAAGTCC-3’ 37 bp 54.05 81.12°C Sense 5’-GTCCAGCCGAATGCGGAGGGCATTCCTTTC-3’ 30 bp 60.00 81.38 Trp472_Ala

Antisense 5’-GAATGCCCTCCGCATTCGGCTGGACAACAGGC-3’ 32 bp 62.50 81.60 Table 2-3: Primers for site directed mutagenesis. Mismatch points are marked in red. The

melting points are calculated from the length (N) and GC content using the formula:

T = 64.9°C + 41°C * (number of G’s and C’s in the primer – 16.4)/N m

(44)

Step Temperature Duration Initial denaturing 95°C 1 min

Denaturing 95°C 1min Annealing 65°C ± 10°C 1 min

Elongation 68°C 8 min

Repeat 20x

Final Elongation 68°C 10 min Hold 4°C

Table 2-4: PCR program for site directed mutagenesis.

Cultivation and expression of AH variants

Large-scale expression was done in 1 l batch cultures using the same anaerobic mineral medium as in the test expressions (Tab. 2-1). 50 mM Na2-fumarate and 0.4% glycerol were used as e^- acceptor and carbon source. Cultures were inoculated with 10 % (v/v) of a stock culture. Cells were grown to OD600 = 1.0 at 37°C. The cultures were then cooled to 25°C and expression of AH was started by addition of 100 µM IPTG. In the case of expression of AH with a chaperone binding site 100 µM TMAO or NaNO3 was also added to induce the corresponding chaperone. After 24 h expression cells were harvested by centrifugation (10,000 x g, 30 min, 4°C). The cell pellet was frozen in liquid N2 and stored at -70°C.

Purification of AH variants

Purification of heterologously expressed AH was achieved as described for the wild type enzyme from P. acetylenicus.

Frozen cells were thawed inside an anaerobic chamber and suspended in 30 ml of 50 mM Tris 200mM NaCl pH 8.0. After addition of DNaseI, MgCl2 and Easy complete protease inhibitor (Roche) cells were disrupted by passing them three times through a French Press at 110 MPa. The French Press was previously stored in the anoxic chamber over night.

(45)

Cell debris and membrane particles were separated from the soluble fraction by ultracentrifugation (100,000 x g, 1:30 h, 4°C). Afterwards, the soluble fraction was subjugated to a two step ammonium sulfate precipitation (2.0 and 3.2 M) After each step precipitated protein was removed by centrifugation (10,000 x g, 30 min, 4°C). Both pellets were suspended in 850 ml 50 mM Tris 200 mM NaCl pH 8.0 and loaded on a Co²⁺ charged Chelating Sepharose Fast Flow column (GE Healthcare). Bound protein was eluted by applying an imidazole and a pH gradient. AH containing fractions were identified by SDS- Page, pooled and concentrated in Amincon Ultra Centrifugal Filter devices 15 (30 kDa cut of, Millipore) to a final volume of 2 ml. The concentrated protein solution was loaded on a SuperDex 200 gelfiltration column (Pharmacia). AH containing fractions were identified by SDS-Page, concentrated to 10 mg/ml, frozen in liquid nitrogen and stored at -70°C until further use.

Crystallization of NarG-AH

Crystals of NarG-AH were grown both by the hanging and the sitting drop vapor diffusion method in an anaerobic chamber (Coy). For initial screening of crystallization conditions screens supplied by Qiagen were used (NeXtal Classics, Classics Lite, MPDs and PACT).

The protein was prepared in 5 mM Hepes/NaOH pH 7.5 at concentrations between 5.0 and 15.0 mg/ml. Buffer exchange was done on a NAP5 column (GE Healthcare). The concentration was adjusted in Vivaspin 500 centrifugal devices (30 kDa cut off, Sartorius Stedim Biotech GmBH). Crystal plates were stored in the anoxic chamber for up to 6 weeks. Crystal formation was checked regularly during this time period. Crystal growth was optimized for conditions in which small crystals formed by changing the concentrations of salt and precipitant, and by changing buffer pH.

For diffraction experiments crystals were transferred to a cryoprotectant solution containing all substances of the reservoir solution and 20% (v/v) MPD. After incubation